Evolution of the PE_PGRS Proteins of Mycobacteria: Are All Equal or Are Some More Equal than Others?

Abstract

PE_PGRS domain proteins represent a family of proteins found in pathogenic and non-pathogenic mycobacteria such as M. smegmatis. This conserved family is characterized by two distinct regions denoted as the variable PGRS domain defined by glycine-rich repeats, and a PE domain consisting of two antiparallel alpha-helices. There are many indications that PE_PGRS proteins are involved in immunopathogenesis and virulence by evading or triggering the host immune response. However, there is not yet any information on their degree of specialization or redundancy. Computational analysis and structural annotation using AlphaFold3 combined with other tools reveals an exceptionally powerful and unprecedented ability to undergo phase separation by the PGRS domain. This suggests that PGRS's glycine-rich, multivalent, low-complexity composition supports phase separation while adopting a structured conformation, contrary to the disordered nature typical of such domains. While previously never reported, the hypothesized role of PGRS in virulence indicates a novel window into the seemingly ubiquitous role of phase separation in cellular compartmentalization and molecular dynamics. This review aims to summarize the current understanding of the PE_PGRS family and its various biological roles in the context of bioinformatic analyses of some interesting representatives of M. marinum that are under control by host sterols. Based on the structural bioinformatics analysis, we discuss future approaches to uncover the mechanistic role of this intriguing family of mycobacterial proteins in both pathogenic and non-pathogenic mycobacteria.

Keywords: AlphaFold3; Mycobacterium tuberculosis; PE_PGRS; host immune response; phase separation.

Figure 1

Neighbor-joining tree of the selection of PE_PGRS proteins with EspB of M. smegmatis included as an outgroup. The tree was based on the MUSCLE alignment and constructed by the neighbor-joining algorithm as implemented in Geneious Prime 2023.2.1 (https://www.geneious.com (accessed on 2 December 2024)). Branch numbers refer to bootstrap values (percentages) based on 1000 replicates. The protein sequences used for constructing this tree are given in Supplemental Table S1 in FASTA format. Mnewor is an abbreviation for M. neworleansens.

Figure 2

A brief overview of PE_PGRS protein activation with TLR2 in the host cell. Both PE_PGRS proteins and TLR2 are located on the membranes. The interaction of PE_PGRS with TLR2 could activate the pathway of NF-κB, leading to the production of cytokines involved in the immune responses. This figure is made by Biorender.

Figure 3

Genetic organization of the M. marinum PE_PGRS1 gene compared to other mycobacteria. Panel (A): The orientation of the orthologs of the PE_PGRS1 gene from M. marinum strain and the two neighboring genes. The gene preceding PE_PGRS1 has been called ucpA based on the annotation of its ortholog in M. avium. The ortholog of this gene in M. tuberculosis has been assigned Rv3549c and its transcription was shown to be induced by. Orthologs of the M. marinum echA20 gene have been annotated as such in many other mycobacteria, but is annotated as CaiD_1 in M. avium. The PE_PGRS1 gene of M. marinum has two copies of very similar orthologues in a large group of related bacteria including M. ulcerans, as shown in the neighbor-joining tree and fast minimum evolution tree generated by fast minimum evolution settings at NCBI: https://blast.ncbi.nlm.nih.gov/Blast.cgi (accessed on 6 December 2024) shown in Figure 1 and Supplemental Figure S1. The PE_PGRS1 genes of M. marinum and M. ulcerans are surrounded by two highly preserved sequence boxes shown in panel (B). Panel (B): Alignment of the highly preserved sequence (HPS) boxes surrounding PE_PGRS1 of M. marinum and M. ulcerans, compared to the HPS boxes in between the orthologues of ucpA and echA20. A BLAST search (word size 16) identified 1087 hits with highly similar sequences in genomes of strains of the genus Mycobacterium (Supplemental Table S2). An additional 132 hits were found with a BLAST+ 2.16.0 search of the M. abscessus HPS box (Supplemental Table S2). The sequence logo is based only on the sequences depicted in panel (B).

Figure 4

Predicted structure of M. marinum PE_PGRS1 made using AlphaFold3 on AlphaFold Server and annotated in PyMol. Prediction values are given in Supplemental Figure S5A. (A): The surface and ribbon representation show the different domains of the protein through having color-coded the different domains with red for the PE region, yellow for the flexible linker between PE and PGRS, cyan for the PGRS domain, and blue for the glycine distribution in the PGRS domain. (B): The inset shows the repetitive motif of double glycine flanking of reactive residues, with a focus on F385 and Y457 on the left, and R271, R323, L320, Q317, and L382 on the right.

Figure 5

Sequence-based phase separation predictions using the PLAAC and PScore algorithms plotted using Python 3.9.0. (A): Bar plots show the PScore per residue for proteins M. marinum PE_PGRS1 and human FUS protein, with the PScore on the Y-axis and residue numbers on the X-axis. (B): PLAAC-generated plots predicting the presence of two prion-like regions (PrD-like) in red. (C): PLAAC-generated plot showing the predicted protein disorder based on the IUPRED3 algorithm.

Figure 6

Predicted interaction between M. marinum PE_PGRS1 and D. rerio TLR2, made using AlphaFold3 on AlphaFold Server and the “find contacts” action in Pymol. The two proteins have been colored cyan and orange, respectively. The interaction interface is strongest and most stable at the PE region, with the inset showing in more detail the two identified polar contacts at PE_PGRS1 residues E47 and A51. Prediction values are shown in Supplemental Figure S4.